fix(VTimePicker): separator size based on density

[Haviles04]

Description

fixes: #22056

Markup:

<template>
  <v-app>
    <v-container>
      <v-row justify="center">
        <v-defaults-provider :defaults="{ global: { density: 'default' } }">
          <v-time-picker format="24hr" use-seconds />
        </v-defaults-provider>
        <v-defaults-provider :defaults="{ global: { density: 'comfortable' } }">
          <v-time-picker format="24hr" use-seconds />
        </v-defaults-provider>
        <v-defaults-provider :defaults="{ global: { density: 'compact' } }">
          <v-time-picker format="24hr" use-seconds />
        </v-defaults-provider>
      </v-row>
    </v-container>
  </v-app>
</template>

[J-Sek]

We need to admit that there is no support for density in VTimePicker. Adding it because of separator is not really helping much. Typically, density is resolved by subtracting (see extracted list of Sass variables that control it). I wish we have at least give it a try to make VTimePicker aligned with all other components.

[Haviles04]

We need to admit that there is no support for density in VTimePicker. Adding it because of separator is not really helping much. Typically, density is resolved by subtracting (see extracted list of Sass variables that control it). I wish we have at least give it a try to make VTimePicker aligned with all other components.

I agree this is a hacky fix, however I thought it was worthwhile looking into because it looks terrible when you have a global density. Another idea would be to just force default density on the buttons

Maybe we can sync up on what density would look like for the time picker and plan out how it would go

[J-Sek]

-4, -8 ? demo

I was trying to align with the spec that changes sizes when using seconds... but maybe there is nothing to worry about. We can make it apply only to the "default" density.

[Haviles04]

-4, -8 ? demo

I was trying to align with the spec that changes sizes when using seconds... but maybe there is nothing to worry about. We can make it apply only to the "default" density.

I think the changes you pushed to the PR look good